(First published in the WAO/FACTOR Newsletter in May 2012)
I just recently read an article in WIRED about Klout that made me want to write more than a tweet about the index so many love, and so many despise. With social media being all the rage, one is not surprised that several companies would want to become THE reference that would help marketers figure out what to do with them. Those companies, at least the few major ones, have adopted the index approach, i.e. that they assign a number that is supposed to represent a level of importance, thus establishing ranks (well, since this is “social”, what better than hierarchy?).
Klout is certainly the most important index on the market (OK, yes, I guess it’s debatable if you favor PeerIndex), with such prevalence that more and more people are taking it into account for a variety of activities (party invitations, job interviews, gifts, etc.). Yes, some people get invited to parties, or even get a job (or rejection, as the WIRED article shows) based on their Klout index. For a good analysis of it, I recommend that you go read that article. Here, I am more interested in what value the index (and any other index of that sort for that matter) has to us online marketers.
Again, as I said elsewhere, what most interests marketers with social media (yeah, yeah, the “conversation” is one too. You really believe it?) is the ability to influence influencers. I also pointed out how flawed I think the concept of influencer is online, or at least how it is determined by social media analytics applications, putting too much weight on audience size in the calculation, missing some fundamental social network dynamics.
From my own experience (and of several people I talked to), one’s activity on Twitter seems to impact Klout a lot. Pausing it, even for a couple of days (the typical length of the weekend) appears to have a quick negative effect on the number. This would mean that, since there’s a premium on simply being active, people with high Klout numbers are in essence people tweeting/posting a lot. Together with being active, audience size, as I said, weighs a lot too. Hard to tell which one is better, though: being very active with a small following, or not doing much, but with a very large audience (which is usually the exclusive realm of celebrities).
Obviously Klout takes into account what people do with influencers’ messages, i.e. whether they’re shared or not, although I don’t know how exactly that sharing fits into the algorithm. Is it the number of sharing actions (likes, RTs, etc.)? A percentage of messages that are shared or a percentage of people sharing over the total following? And how much not being shared a lot (the heart of social in social media) would one’s score suffer if one’s accounts had millions of followers? And we need to ask: does sharing someone else’s messages any real indication that one is influenced by that person? If influence is the ability to make someone changes opinions, or adopt a new behavior, then I think not.
Besides the complexity of multi-factor indexes, and how hard their variations are to understand, we have a scale problem. Is someone at 45 half as influential as someone with 90? Can one even say that about influence, that this person is “three times” more influential than this person?
I think we are not there yet, as with a lot of other social media measurement schemes. And as marketers, we should be very, very careful when we want to invest our money based on such fuzzy things as social media indexes.
A final note: at the time of writing this, my Klout just got down to 45 from 47 in a week. You will find me much more on Twitter from now on, slavering to get to 50, and wait for those nice perks.